Circulation 1969 WITTE 723 33

RADER, NORMAN LEVINE and ERNEST S. BREEDMARLYS HEARST WITTE, ALLAN E. DUMONT, ROY H. CLAUSS, BERTHA

DrainageLymph Circulation in Congestive Heart Failure: Effect of External Thoracic Duct

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1969 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.39.6.723

1969;39:723-733Circulation.

http://circ.ahajournals.org/content/39/6/723located on the World Wide Web at:

The online version of this article, along with updated information and services, is

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. and Rights Question and Answer

Permissionsthe Web page under Services. Further information about this process is available in thewhich permission is being requested is located, click Request Permissions in the middle column ofClearance Center, not the Editorial Office. Once the online version of the published article for

can be obtained via RightsLink, a service of the CopyrightCirculationoriginally published in Requests for permissions to reproduce figures, tables, or portions of articlesPermissions:

by guest on October 8, 2014http://circ.ahajournals.org/Downloaded from by guest on October 8, 2014http://circ.ahajournals.org/Downloaded from

http://circ.ahajournals.org/content/39/6/723

http://www.ahajournals.org/site/rights/

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/

http://circ.ahajournals.org/


Lymph Circulation in Congestive Heart Failure

Effect of External Thoracic Duct Drainage

By MARLYS HEARST WIrTE, M.D., ALLAN E. DUMONT, M.D.,

RoY H. CLAUSS, M.D., BERTHA RADER, M.D., NORMAN LEVINE, B.S.,

AND ERNEST S. BREED, M.D.

SUMMARYThe lymphatic circulation was evaluated after cannulation of the cervical thoracic

duct in 12 patients with severe intractable congestive heart failure. After venting thedistended duct, lymph flowed rapidly under increased pressure, and signs and symptomsof circulatory congestion were dramatically relieved. Dyspnea, orthopnea, anorexia,abdominal discomfort, distended neck veins, hepatomegaly, peripheral and scrotaledema, and ascites diminished or disappeared. Central venous pressure fell and arm-

to-tongue circulation time decreased. In some patients arterial pressure, amplitude ofperipheral pulses, central venous Po2, and urinary output rose further suggestingenhanced cardiac output. Considered together with previous observations in experimentalanimals, the findings indicate that elevated systemic venous pressure greatly increasesformation of thoracic duct lymph but at the same time impedes the return of lymphinto the great veins. While therapeutic drainage of lymph from the thoracic duct inpatients with cardiac failure is limited, the capacity of the lymphatic system totransport excess capillary filtrate back to the blood stream constitutes a major controlmechanism regulating the manifestations of this disorder.

Additional Indexing Words:Central venous pressure Circulation tin

C ONTROVERSY concerning the patho-genesis of edema in congestive heart

failure has centered around the manner in

From the Departments of Medicine and Surgery,New York University School of Medicine, New York,and the Department of Medicine (CardiologyDivision), Washington University School of Medicine,St. Louis, Missouri.

This research was supported by Grants HE-09073,HE-11904, HE-11034, and AM-20875, from theU. S. Public Health Service, and funds from theAmerican Heart Association, St. Louis and MissouriHeart Associations, National Science Foundation, JohnA. Hartford Foundation, and the Institute of MedicalEducation and Research of St. Louis City Hospital.Some of these patients were studied in the U. S.Public Health Service-supported clinical research unitat Bellevue Hospital, New York.

Dr. Witte is an Established Investigator of theAmerican Heart Association and Dr. Dumont is arecipient of a U. S. Public Health Service CareerResearch Development Award.

Circulation, Volume XXXIX, June 1969

ie Metabolic balance studies

which decreased cardiac output and increasedsystemic venous pressure promote formationof excess capillary filtrate.1 2 According to oneview, failure of the heart leads directly to arise in venous pressure which is reflected in arise in capillary pressure, increased capillaryfiltration, and extravascular fluid accumula-tion. Salt and water are retained by thekidneys in an effort to restore circulatingblood volume. The alternative view holds thatreduction in cardiac output itself stimulatessalt and water retention by the kidney.Hypervolemia follows and leads to increasedvenous and capillary pressure, increasedcapillary filtration, and accumulation ofextravascular fluid which is aggravated byanoxic damage to the capillary endothelium.

Presented in part at the annual meeting of theAmerican Heart Association, Bal Harbour, Florida,November 24, 1968.

723

by guest on October 8, 2014http://circ.ahajournals.org/Downloaded from


724 WITTE ET AL.

Table 1

Data from 12 Patients with Congestive Heart Failure in Whomn External Thoracic Duct Drainage WVas Performed

Age (yr)and Sex Diagnosis and clinical findings

1. R.J. 47 F RHD: MS, TS, TR, PH, AF, PVC, digitalistoxicity, cardiac cirrhosis IV-E

cvpressure(cm H 20)

Previoustreatment

40 d, m, t

Thoracic duct lymph

Flow Pressure(ml/min) (cmr H,20)

3 17

2. A.D. 33 M RHD: MS, PAT, PVC, bigeminy, 7 yearspost MC, IV-E

3. M.B. 45 F RHD: MS, PH (PPA mean 75) AF, IV-E,CI 1.4 L (fixed), serum Na 128, K 5.9

4. C.B. 51 M Marie Strumpell spondylitis: associated AR,1 yr post-aortic valvuloplasty andaortoplasty, pulmonary emboli, intractableanginaIV-E

5. M.C. 73 F RHD: MS, MR, TR, cardiac cirrhosis, IV-Eobtunded. serum Na 127, K 5.6

6. E.B. 50 F RHD: MS, MR, TR, 3 years post MC, IV-E

7. R.Z. 41 M RHD: AS, AR, MS, MR SBE (active),cerebral and pulmonary emboli, IV-Eobtunded, serum Na 122, K 3.9

8. E.M. 58 M HHD, ASHD: MI, ruptured IV septum,pulmonary emboli, IV-E,serum Na 126, K 3.6

9. T.M. 47 F CHD: IASD + ? MS, IV-E,anuria, severe hypotension, no palpablepulses. Serum Na 112, K 6.3

10. V.T. 49 F RHD: MS, 9 years post MC, PH(PPA 110/30), AF, cardiac cirrhosis, IV-Ehypotension, severe oliguriaserum Na 118, K 6.8

11. M.V. 47 F UHD, extreme obesity, PVC, digitalistoxicity, acute and chronic respiratoryinsufficiency with coma, IV-E, anuria

12. J.B. 45 M Idiopathic cardiomyopathy,pulmonary emboli, IV-Egangrene RLE

28 d, a, t

24 d, adl, m, t,paracentesis,phlebotomy

26 d, a, acl, am,m, t,paracentesis,antithyroid drugs,morphine, abt

27 d, acl, m, t,Southey tubes

25 d, a, adl, m, t

27 d, a, adl, m, t,Southey tubes,abt, aeg

23 d, a, adl, am, dx,m, t,thoracentesis, aeg

37 d, a, m, tabt, aeg

30 d, a, am, m, t,morphine,Kayexalate

90 d, a, adl, m, t,positive pressureventilation

18 d, a, m, t,phlebotomy,aeg

Abbreviations: AF = atrial fibrillation; AR = aortic regurgitation; AS = aortic stenosis; CI = cardiac index (L/min/M2); CV = central venous; HD = heart disease: AS = arteriosclerotic, C = congenital, H = hypertensive, R =rheumatic, U = unknown; IASD = interatrial septal defect; IVSD = interventricular septal defect; IV-E = func-tional-therapeutic classification; LA = left atrium; LV = left ventricle; MC = mitral commissurotomy; MI = myo-

cardial infarction; MR = mitral regurgitation; MS = mitral stenosis; PPA = pulmonary artery pressure (mmnHg);PAT = paroxysmal atrial tachyeardia; PH = pulmonary hypertension; PM = postmortem; PV = plasma volume;PVC = premature ventricular contractions; RLE = right lower extremity; RLL = right lower lobe; RV = right ven-


Patient

7 28

5 30

11 80

9

12.5

4

23

43

2 26

30 30

3 20

3 43

4.5 19



LYMPH CIRCULATION IN HEART FAILURE

Tlioracic duct lymiph

Total protein 45-min BSPg% (% plasmia) nmg% (% plasma)

3.3 (82%)

1.4 (21 O)

6.0 (75%c)

Duration Weightdrainage loss(days) (kg) Additional studies

7 10.9 Figs. 3, 4 (top), 5

7 12.3 Figs. 3, 4 (top), 5

7 3.9 Figs. 3, 4 (top), 5PV 3.8 A1.4L

1.7 (31%) 8 25.0 Figs. 2-4 (top),5, 7

2.6 (39%) 3 10.9 Figs. 2-4 (top),5, 7

6.0 (85%) 2 0.45 Figs. 3-5PV 3.9 -* 2.6L

2.0 (33%) 1.05 (17%) 3 8.6 Figs. 3-7urine aldosterone

3.8 (52%) 1 1.09 Figs. 3, 4 (top), 5

3.3 (53%) 5 15.5 Figs. 2-5, 7

1.6 (35%) 1.46 (34%) 5 7.3 Figs. 3, 4 (top), 5-7t 1/2 plasmaaldosterone

2.8 (32%) 0.75 (22%) 5 12.7 Figs. 3, 4 (top),5-7

0.7 (11%) 1.75 (25%) 3 5.0 Figs. 3, 4 (top),5, 7, 9

Course

2 wks later: CI 2.85L, PPA mean 48.Discharged after wt loss of +30 kg on d,a,e, m, t. Doing well +1 yr on d, a, m, t.

Cannula removed. Sudden death duringPAT 1 day later.

Discharged on d, a, m, t. Severe CHF for2 yr. PM: mitral valve diameter 2 mm,obliterative thrombus LA, multiplepulmonary emboli.

Lymph clot in cannula. Died in pulmonaryedema. PM: valvuloplasty incompletewith torn sutures, MI, bronchopneumonia

Cannula fell out. Discharged on d, a, and m.Doing well 6 mo later.

Surgery: mitral and tricuspid annuloplasty.Died 1 day PO.

Discharged on d. Died suddenly at home2 mo later.

Surgery: IVSD repaired. Died PO.

Pulses became palpable, extremities warm;CV oxygenation T . Cannula removed.Died suddenly 2 days later.

Cannula clamped. Hypotension reappeared.Died 1 day later.

Awakened from coma. Urine output roseto normal. CV oxygenation T. Day 5:died after massive aspiration of food.PM: UHD.

Surgery: RLE amputation. Died 2 dayslater. PM: thrombi RV, LV, and commoniliac artery; RLL pneumonia and infarction


725

tricle; SBE = subacute bacterial endocarditis; TR = tricuspid regurgitation; TS = tricuspid stenosis.Abbreviations concerning treatment: a = Aldactone (spironolactone); abt = antibiotic; acg = anticoagulant; acl =

ammonium chloride; am = aminophylline; d = digitalis; dx = Diamox (acetazolamide); e = ethacrynic acid; m =meralluride (Mercuhydrin); t = thiazide.



2WITTE ET AL.

The controversy remains unresolved, for whileeither theory explains certain clinical andexperimental observations, neither one iscompatible wvith all.

Fluid that leaves the intravascular compart-ment normally returns via venous andlvmphatic pathways. However, once the heartfails as a pump, reabsorption is graduallyimpeded, the vascular space becomes over-loaded, and fluid shifts into tissues. As venouscongestion becomes intense, edema accumu-lates at a rate limited only by lymph flow andcompliance of the interstitial space.3

Starling was the first to suggest thatderangements in the formation and transportof lvmph bear directly on the pathogenesis ofcirculatory congestion in heart disease. Heconceded, however, that his views, based inpart on direct observations in a single dogwith spontaneous heart failure,4 were largely'<matters of surmise." In the present study, thelvmph circulation was evaluated in 12 patientsNwith severe congestive heart failure. Flow,pressure, and composition of thoracic ductlymph were determined, and the effects ofexternal lymph drainage on the manifestationsof heart failure were examined. Theseobservations extend and amplify earlierfindings.5

MethodsTwelve patients with severe congestive heart

failure of several years' duration were studied.Pertinent clinical data are presented in table 1.Marked dyspnea, orthopnea, anorexia, distendedneck veins that filled from below, enlarged tenderliver, ascites, and peripheral edema were presentin all patients. Treatment with rest in bed, low-salt diet, digitalis, a variety of diuretic drugs, andmechanical removal of intracavitary or peripheralfluid had failed, and most patients were moribundwith associated severe renal and electrolvtedisturbances. In three patients closed cardiacsurgery and in one patient open heart surgery hadbeen performed previously. Functional-therapeu-tic classification was IV-E in all patients.

Under local anesthesia the cervical portion ofthe thoracic duct was cannulated with polyethyl-ene tubing and lymph was collected by gravitydrainage. Flow rate (ml/min) and end pressure(cm H11O above the right atrium) weredetermined. Body weight, oral and intravenousintake of fluid, and output of lymph and urine

were measured daily. Central venous pressurewas monitored manometrically via a catheter inthe superior vena cava or right atrium. Externallymph drainage was continued for 1 to 8 days.Based on vital signs, central venous pressure,urinary output, and frequent hematocrit readings,lymph losses were partially replaced by increasingoral intake of salt-containing fluids or intravenousinfusion of balanced salt solutions, in someinstances wvith added serum albumin. No diureticdrugs were administered, but the dailymaintenance dosage of digitalis (0.1 to 0.2 g ofdigitalis leaf or 0.25 to 0.5 mg of digoxin) wascontinued except in two patients in whomdigitalis toxicity was suspected. After conclusionof the drainage period, the cannula was removedat the bedside and a pressure dressing applied.

Additional studies were performed as indicatedin table 1 and figures 1 to 7: total protein contentof lymph and plasma (12 patients), serialdeterminations of arm-to-tongue circulation timedehydrocholic acid (Decholin; three patients),serial levels of renin in plasma and lymph(modified method of Hellmer and Judson;6' 7three patients), serial determinations of urinaryaldosterone excretion (double isotope dilutionderivative assav;S one patient), disappearance rateof injected 3H-d-aldosterone (one patient),lymph and plasma BSP content 453 min afterintravenous injection of 5 mg/kg (four patients),and serial plasma volume determinations (Evansblue dye-dilution method; two patients).The thoracic duct was also cannulated in four

patients with "compensated" heart disease (novisible edema) prior to corrective cardiac surgery

TDL FLOW TDL PRESSURE

Figure 1Mean flow, end pressure, and total protein content ofthoracic duct lymph (TDL) in patients with heartfailure (compensated and decompensated) and incontrol patients.

Circulation, Volume XXXlX, June 1969

7i 6




gV

BEFORE AFTER

Figure 2

Dimtrinution int distention of neck veiens (top), ge

edema (bottom) after drainage of lymph fromcongestive failure.

(valve replacement [three patients] and closureof patent ductus arteriosus [one patient]) and in43 control subjects" during the course of scalenenode biopsy for diagnosis of a variety of disordersunassociated with liver or heart disease.

Results

In patients with congestive heart failure, thethoracic duct was enlarged to a diameter of 5to 12 mm (normal, about 2 mm). Thoracicduct lymph flow measured promptly afterCirculation, Volume XXXIX, June 1969

!nital and thigh edema (miiddle), and pedolthe thoracic duct in three patients with

placement of the cannula was greatly elevated(mean, 7.8 ml/min; range 2 to 30 ml/min;normal, 0.82 0.27). Lymph end pressure

was greatly increased (mean, 32.4 cm H20;range, 17 to 80 cm H2O; normal, 11.6+ 6.6)and closely approximated central venous

pressure (imean, 32.9 cm H20) in mostpatients (table and fig. 1).There was no correlation between the

magnitude of the initial thoracic duct lymph

CASE4

CASE5

727

igiL..:: I



WITTE ET AL.

444

27lFZ lallnsght lossED 7o:l yasYphoutputm Total auMe output

21

2 3 4 5 6 7 8 9 10 l 12CASE #

Figure 3Total weight loss as related to lymph output and urineoutput during entire period of lymph drainage inpatients with congestive failure. Lymph losses werepartially replaced by oral and intravenous administra-tion of salt-containing fluids.

flow and the level of initial central venouspressure although individual patients general-ly exhibited a decline in lymph flow with theearly sharp fall in central venous pressure.Lymph flow (mean, 1.2 ml/min; range, 0.75 to2.0 ml/min) and end pressure (mean, 10 cm)in four patients with "compensated" heart

CENTRALVENOUSPRESSURE(cm H20)

CIRCULATIONTIME(seconds)

Figure 4

Decrease in central venous pressure and arm-to-tongue circulation time after lymph drainage in pa-

tients with congestive failure (numerals represent case

numbers).

3000O

-O

- 2000E

/2

ocOpen HeatSurgery Performed

0Z5-

55

BEFORE DURING

Figure 5AFTER

Mean daily urine output before, during, and afterconclusion of lymph drainage in patients with con-gestifie failure (numerals represent case numbers).

disease did not differ significantly from normalvalues (fig. 1).

Total protein content in thoracic ductlymph (table 1 and fig. 1) was decreased to46% of plasma (P <0.01; range, 11 to 82%)while values in patients with compensatedheart disease (mean, 70%; range, 57 to 79%)did not differ significantly from controlsubjects (72% + 11%). Lymph BSP at 45 min

DAYS

Figure 6

Changes in circulating renin levels (expressed asnanograms of angiotensin per ml) in thoracic ductlymph and systemic venous blood during thoracicduct lymph drainage in patients with congestive failure(numerals represent case numbers).


"728

7

I



LYMPH CIRCULATION IN HEART FAILURE 729CASE 4 CASE 5

DwCommula Du~~ ~ ~~~~~~~~~~~~~~~~~~ctCvo4Io1Ca,,n.lat-o-Clotted ~~~~~~~~~~~~~~~~~~~~~~Ca.nuo6i,onClotted

V" o.S ~~~~~~~~~~~~~60 Cent, o300 -. 40

(8Hn K --. 40 .BUN)8g%) 0l 788o,80 200. -BUN (mg%)ioo 20 Dcr~~~~~~~~~~~~~~~~~~bO 20

0 P1115s, ~~~~~~10 0'0000- L ~~~~~~~~~~~~~~~~~~~~~~~~No 0120

112010 (kg)

800 4000+ CL 65I 800 40Dl8I 00 8

I ~~~~60 oTo8(4002000 24001.______________P 2 004

-0-0u55 00~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~5

1 3 4 -0 LU5~~~~~~~~~00800 I50()

DAY 2 3 4 56C01416 0088802. DAY~~~~~~~~~~~~~~~~~~~~~A

CASE 10 CASE E11Tho~~~~~oc'c82010.142410 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ hoaccuc

Cannulot.on Ren'av.d~~~~~~~~~~~~~~~~~~~~~~onuotoConnu001Asllo8

rCoeci Dut o300al 00.8111000 00Pv.ss.re 200 0116)1%)~~~~~~~Remve 06,8 00 8

(001120)' - 8BU (.T1h4%)i 3T10,0 600 6

00 200.01 '40UNBNmgglVenous 200

J _ '40 (mm KO)~~~1eg , 40- 40--Pressure - -iomet All 0 0~~~~~~~~4-8-loo 20 (r,, Eq)~~0 (811n2) 20

0 0~~~~~~~~~~~~~~~~~~

40000 SWo gh LW60gWe.LhWe6000(kg) kg

5000~~~~~~~~~~~4"43000

55 400 40~~~80 40

2000o naeOtuL 50 H20 U~~~~~~~~~50202000 4

1 2 3 ~~~~45 05_ L_1 2 3 4 5 1 ~~~~~~~~~~~~~~~~~~~~~~~~1 2 3 4 5 6

Cnfl DAY LCly ph UNTie RONTO DAY (HiTL

Metaolaic balace tudes urigtoai utlmhdangei i ainswt hoiconestveheattailre Symptm n in fcicltr ogsindakdydmnse

exrto eandlweelssofbloodMura)nitroent(rNa arelu ddntcorltwithur chngsinurnryoupt.Alhug( lmhIlo enral declied i0thefirs fe Hour

it fell spontaneously and in case 4 where the tip of th~~~Pessdraining canuawa 4eeatd




WITTE ET AL.

xvas 25% (range, 17 to 34%) of thesimultaneous plasma level (table 1). Nosignificant retention of BSP was observed at45 min in plasma of control subjects"' and intwo patients with compensated heart disease.

XVithin a few hours after thoracic ductcannulation, important changes were noted inthe clinical status of each patient, and thesechanges persisted throughout the entire periodof lymph drainage. Dyspnea and orthopneadiminished or disappeared, and anorexia andabdominal discomfort were relieved. Distend-ed neck veins, bepatomegaly, peripheral andscrotal edema, and ascites also diminished ordisappeared (fig. 2). Body weight decreased

VENACAVA THORACIC DUCT

PERIPHERAL SPLANCHNICBED BED

NORMAL

OUTFLOWBLOCK

14

O EAo

(ITOOUTFLWEDEMA BLOCK

HEART HEPATICFAILURE CIRRHOSIS

Figure 8

Schematic comparison of thoracic duct lymph forma-tion and retuirn to the venous system in heart failureand hepatic cirrhosis. In heart failure, generalizedvenotus hypertension promotes excess lymph (o) forma-tion in both peripheral and splanchnic tissues. At thelymphatic-venous junction in the neck, local resistanceand increased central venous pressure impede lymphreturn, which leads to lymphatic dilatation and hy-pertension and edema (a). In hepatic cirrhosis, venoushypertension is confined to the splanchnic bed andincreased volumes of splanchnic lymph are returnedvia the thoracic duct to normotensive veins. Whenthe lymphatic system is overloaded, excess capillaryfiltrate leaks into the peritoneal cavity to form ascites(-).

(mean, 9.5 kg; range, 0.5 to 25 kg) andcorresponded to unreplaced lymph losses(table 1, fig. 3).Central venous pressure fell from a mean of

32.9 to 14.0 cm 112O (table 1, figs. 4 top and7). Lymph flox rates fell in most patients asvenous pressure decreased acutely, butabnormally rapid flow persisted even aftervenous pressure returned to normal. Arm-to-tongue circulation time decreased (fig. 4bottom). Urinary output increased in fourpatients who were oliguric or anuric prior tothe procedure (figs. 5 and 7). Plasma volumedecreased in tvo patients (table 1).Renin levels in lymph and plasma did not

change consistently and remained elevated(fig. 6). In case 7 (fig. 7), urinary

Figure 9

Case 12. Postmortem dissection of junction betweenthe thoracic duct (TD) and subclavian vein (SV). Notemarkedly dilated thoracic duct (1 cm diameter) prox-imal to narrow channel of entry (arrow) into subelavi-an vein. (Dissection by Dr. Charles L. Witte.)


7:30




aldosterone excretion ( 154 ug/day before and380 ,ug/day after drainage) and renin levels inlymph and plasma remained elevated despitemarked clinical improvement. In case 10 (fig.7), the biological half life of aldosteroneremained prolonged (55 min compared with50 min preoperatively; normal, 20 to 40min)" despite a decrease in venous pressureand diminution in edema.Four patients were subsequently discharged

with striking improvement in their clinicalstatus (table 1). Three other patientsunderwent surgery immediately after lymphdrainage. Two patients died during lymphdrainage and the remaining three died withinone week following removal of the cannula.

DiscussionThe partition of extracellular fluid between

the vascular and interstitial fluid compart-ments is determined by the net effect ofcapillary filtration, plasma leakage throughlarge pores, venous reabsorption, and lymphflow. Generally, as venous pressure is raised,filtration of solutes increases more than bulkloss of protein-rich plasma.3 As a result, tissueprotein content and thus oncotic pressuredecrease and these changes initially counter-balance the rise in capillary hydrostaticpressure. As venous pressure rises still further,this compensatory mechanism is insufficient,and filtration is favored even at the venousend of the capillary. At this point, excessinterstitial fluid accumulates and only in-creased lymph flow can remove it.3

In patients with congestive failure, theincreased tendency for fluid to leave thevascular space and the great capacity of thelymphatic system to transport excess tissuefluid is reflected in the increased flow oflymph under high pressure from thecannulated thoracic duct. Normally, thoracicduct lymph arises almost entirely from theliver and extrahepatic portal bed.'2 Liversinusoids are freely permeable to plasmaprotein, and the absence of an oncoticgradient between plasma and hepatic intersti-tial fluid means that small increases inhydrostatic pressure produce large increases information of liver lymph, which remains highCirculation, Volume XXXIX, June 1969

in protein.'13 14 On the other hand, extrahe-patic portal capillaries are less permeable, andas venous pressure rises, the protein content ofintestinal lymph falls.14 In chronic severecongestive failure, thoracic duct lymphprotein is usually greatly diminished, a findingwhich indicates that excess lymph originatesnot only from the liver but also in large partfrom less permeable extrahepatic portal andperipheral capillaries which form a progres-sively more dilute filtrate as venous pressurerises. Similarly, 45-min BSP content ofthoracic duct lymph (relative to plasma), amarker of liver lymph in the thoracicduct, 10, 15 is also low, as it is in patients withadvanced hepatic cirrhosis in whom extrahe-patic portal hypertension predominates.'0Although lymph flows at an accelerated rate

when the thoracic duct is cannulated, thepresence of edema prior to this maneuvermeans that flow in the unvented lymphaticsystem has not kept pace with the rate oflymph formation'6' 17 (fig. 8). In 1937,McMaster18 observed that colored dyesinjected subcutaneously into the edematouslegs of patients with cardiac failure enteredwidely dilated, richly intercommunicatinglymphatic channels but consistently failed toshow "streamer" formation or forward flow.Instead, coloring matter pooled in lymphatics,extravasated into tissues, and even flowedretrograde. Great dilatation of the thoracicduct and its tributaries, ballooning of the ductas it approaches the relatively narrow junctionof the thoracic duct with the jugular vein (fig.9), equilibration of thoracic duct end pressureand central venous pressure, and sluggish to-and-fro motion of venous blood refluxingthrough the cervical junction further attest togeneralized impairment to lymph flow inheart failure.At least two important factors limit the flow

of lymph into the venous system: localresistance at the thoracic duct-jugular veinjunction'9' 20 and central venous pressureitself. In patients and experimental animalswith right heart failure, thoracic duct lymphflow, measured by several different methods,is impeded by a high outlet pressure, although

731



7WITTE ET AL.

flow into the vein may be somewhat greaterthan normal for a given pressure.21 24 Thoracicduct end pressure, the net force propellinglymph onward to the great veins, is alsoincreased,24.2 but generally to levels nohigher than corresponding central venouspressure, thus providing little or no gradientfor lymph flow.

Cannulation of the distended thoracic ductresults in a copious flow of lymph andmobilizes excess extravascular fluid. As lymphis drained externally and only partiallyreplaced, the signs and symptoms ofcirculatory congestion diminish or disappear.The lymphatic system is again able to siphonoff excess fluid from the venous system, andcentral venous pressure falls. Arm-to-tonguecirculation time decreases, and this decreasesuggests improvement in cardiac output,supported in some patients by a rise in arterialblood pressure, increased amplitude ofperipheral pulses, improvement in centralvenous oxygenation, and increased urineoutput.)6 However, urinary sodium excretionremains low and the renin-aldosterone systemis not suppressed.

Furthermore, the manifestations of circula-tory congestion can be relieved even withoutexternal lymph losses if lymph return isincreased to match lymph production. Forexample, vhen the thoracic duct is anasto-mosed to a normotensive pulmonary vein indogs with isolated right-heart failure frompulmonic stenosis and tricuspid insufficiency,24sequestered lymph behind the right heart iseffectively diverted to the left heart.Reduction in central venous pressure, natriu-resis, diuresis, and loss of ascites follow.Similarly, in hepatic cirrhosis (fig. 8), anotherdisorder characterized by increased produc-tion of thoracic duct lymph,10 19 edema in theform of ascites is not always found. Here,venous hypertension is confined to thesplanchnic bed, and as long as excess lymphdrains into normotensive systemic veins asrapidly as it is formed in the splanchnic bed,ascites does not appear.27 Moreover, in dogswith ascites secondary to constriction of thesupradiaphragmatic inferior vena cava or

hepatic veins, either enlargement of thethoracic duct-jugular vein junction28 oranastomosis of the thoracic duct to intrathorac-ic systemic veins of lower pressure29 30 leadsto increased lymph flow in the thoracic ductand loss of ascites.

In patients with heart disease, whencapillary filtrate forms more rapidly than itreturns, excess interstitial fluid accumulates,and effective arterial volume decreases. Acomplex sequence of events is set into motionleading to retention of salt and water, whichtends to restore this "lost" volume, but at thesame time promotes further capillary filtra-tion.27 The capacity of the lymphatic systemto transport excess filtrate back to the bloodstream plays a key role in regulating themanifestations of congestive heart failure.

AcknowledgmentWe are indebted to Drs. John R. Smith and Charles

L. Witte for help in preparing the manuscript, to Dr.Gabor Kaley for performing the renin assays, and toMiss Amalia Martelli for technical assistance.

References1. PAINE, R., AND SMITH, J. R.: The mechanism of

heart failure. Amer J Med 6: 84, 1949.2. BURCH, G. E., AND RAY, C. T.: Consideration of

the mechanism of congestive heart failure.Amer Heart J 43: 918, 1951.

3. WIEDERHIELM, C. A.: Dynamics of transcapil-lary fluid exchange. In Biological Interfaces:Flows and Exchanges, Proceedings of a Sym-posium sponsored by the New York HeartAssociation, edited by F. P. Chinard. Boston,Little, Brown & Co., 1968, p. 29.

4. BOLTON, C., AND STARLING, E. H.: Note on theblood pressure and lymph flow in a case ofheart disease in a dog. Heart 1: 292, 1910.

5. DUMONT, A. E., CLAUSS, R. H., REED, G., ANDTICE, D.: Lymph drainage in patients withcongestive heart failure. New Eng J Med 269:949, 1963.

6. HELLMER, O., AND JUDSON, W. E.: Quantitativedetermination of renin in the plasma of patientswith arterial hypertension. Circulation 27:1050, 1963.

7. KALEY, G., WITTE, M. H., AND LEVINE, N.:Presence of renin in thoracic duct lymph ofman. (Abstr.) Fed Proc 24: 525, 1965.

8. KLIMAN, B., AND PETERSON, R.: Double isotopederivative assay for aldosterone. J Biol Chem235: 639, 1960.

9. DUMONT, A. E., AND WITTE, M. H.: Clinicalusefulness of thoracic duct cannulation. In


732




Advances in Internal Medicine, vol. 15, editedby G. Stollerman. Chicago, Year Book MedicalPublishers, 1968.

10. WITTE, M. H., DUMONT, A. E., COLE, W. R.,WITTE, C. L., AND KINTNER, K.: Lymphcirculation in hepatic cirrhosis: Effect of port-acaval shunt. Ann Intern Med 70: 303, 1969.

11. WITTE, M. H., LEVINE, N., AND DUMONT, A. E.:Exchange of aldosterone between plasma andthoracic duct lymph. J Clin Endocr 27: 1730,1967.

12. YOFFEY, J. M., AND COURTICE, F. C.: Lymphatics,Lymph and Lymphoid Tissue, ed. 2. Cam-bridge Harvard University Press, 1965, p. 143.

13. STARLING, E. H.: Influence of mechanical factorson lymph production. J Physiol 16: 224,1894.

14. WITTE, C. L., WITTE, M. H., DumIONT, A. E.,FmST, J., AND COLE, W. R.: Lymph protein inhepatic cirrhosis and experimental hepatic andportal venous hypertension. Ann Surg 168:567, 1968.

15. WITTE, M. H., DUJMONT, A. E., LEVINE, N., ANDCoLE, W. R.: Patterns of distribution ofsulfobromophthalein in lymph and bloodduring obstruction to bile flow. Amer J Surg115: 69, 1968.

16. RUSZNYAK, I., FOLDI, M., AND SZABO, G.:Lymphatics and Lymph Circulation. NewYork, Pergamon Press, Ltd. 1960, p. 561.

17. WITTE, M. H., AND COLE, W. R.: Failure oflymph flow in man and experimental animals:Reversal by surgically constructed lymphatic-venous shunts. In Progress in Lymphology,edited by A. Riittimann. Stuttgart, GeorgThieme Verlag, 1967, p. 376.

18. MCMASTER, P. D.: Lymphatics and lymph flowin the edematous skin of human beings withcardiac and renal disease. J Exp Med 65: 323,1937.

19. DUMONT, A. E., AND MULHOLLAND, J. H.:Hepatic lymph in cirrhosis. In Progress in LiverDisease, vol. 2, edited by H. Popper and F.

Schaffner. New York and London, Grune &Stratton, Inc., 1965, p. 421.

20. BRADHAM, R. R., AND TAKARO, T.: Nature andsignificance of the thoracic duct-subclavianvein junction. Surgery 64: 643, 1968.

21. FOLDI, M., AND PAPP, M.: Role of lymphcirculation in congestive heart failure. Jap CircJ 25: 703, 1961.

22. WEGRIA, R., ET AL.: Effect of systemic venouspressure on drainage of lymph from thoracicduct. Amer J Physiol 204: 284, 1963.

23. SZABO, G., AND MAGYAR, Z.: Effect of increasedsystemic venous pressure on lymph pressureand flow. Amer J Physiol 212: 1469, 1967.

24. COLE, W. R., WInTE, M. H., KASH, S. L.,RODGER, M. R., BLEISCH, V. R., ANDMUELHEIMS, G. H.: Thoracic duct topulmonary vein shunt in the treatment ofexperimental right-heart failure. Circulation36: 539, 1967.

25. BLALOCK, A., AND BURNELL, C. S.: Thoracic ductlymph pressure in concretio cordis: Experi-mental study. J Lab Clin Med 21: 296,1935.

26. EICHNA, L.: Circulatory congestion and heartfailure. Circulation 22: 864, 1960.

27. WInTE, M. H., ET AL.: Lymphatic circulation indisorders of salt and water balance resultingfrom hepatic venous oufflow block. Proceed-ings of the Second Intemational Congress ofLymphology. In press.

28. BELAN, A., AND VOSMIK, J.: The abdominal andthoracic duct in experimental portal hyperten-sion, Op. cit.27

29. ZoTTi, E., LESAGE, A., BRADHAM, R., NIGNONE,R., SEALY, W., AND YOUNG, W., JR.: Preventionand treatment of experimentally inducedascites in dogs by thoracic duct to vein shunt.Surgery 60: 28, 1966.

30. COLE, W. R., WITTE, M. H., MALEY, E.,BLEISCH, V., AND ABERNATHY, R. E.: Alymphatic portacaval shunt for the treatment ofexperimental ascites. In Progress in Lympholo-gy, op.cit.17

Circulation, VolUme XXXIX, June 1969

733



Documents

Circulation 1969 WITTE 723 33